Rate-Dependent Conduction Block of the Crista Terminalis in Patients With Typical Atrial Flutter

Transcription

1 Rate-Dependent Conduction Block of the Crista Terminalis in Patients With Typical Atrial Flutter Influence on Evaluation of Cavotricuspid Isthmus Conduction Block Angel Arenal, MD; Jesus Almendral, MD; Jesus M. Alday, MD; Julian Villacastín, MD; Jose M. Ormaetxe, MD; Jose Luis M. Sande, MD; Nicasio Perez-Castellano, MD; Sergio Gonzalez, MD; Mercedes Ortiz, PhD; Juan Luis Delcán, MD Background The crista terminalis (CT) has been identified as the posterior boundary of typical atrial flutter (AFL) in the lateral wall (LW) of the right atrium (RA). To study conduction properties across the CT, rapid pacing was performed at both sides of the CT after bidirectional conduction block was achieved in the cavotricuspid isthmus by radiofrequency catheter ablation. Methods and Results In 22 patients (aged 61 7 years) with AFL (cycle length, ms), CT was identified during AFL by double electrograms recorded between the LW and posterior wall (PW). After the ablation procedure, decremental pacing trains were delivered from 600 ms to 2-to-1 local capture at the LW and PW or coronary sinus ostium (CSO). At least 5 bipolar electrograms were recorded along the CT from the high to the low atrium next to the inferior vena cava. No double electrograms were recorded during sinus rhythm in that area. Complete transversal conduction block all along the CT (detected by the appearance of double electrograms at all recording sites and craniocaudal activation sequence on the side opposite to the pacing site) was observed in all patients during pacing from the PW or CSO (cycle length, ms), but it was fixed in only 4 patients. During pacing from the LW, complete block appeared at a shorter pacing cycle length ( ms; P 0.01) and was fixed in 2 patients. In 3 patients, complete block was not achieved. Conclusions These data suggest the presence of rate-dependent transversal conduction block at the crista terminalis in patients with typical AFL. Block is usually observed at longer pacing cycle lengths with PW pacing than with LW pacing. This difference may be a critical determinant of the counterclockwise rotation of typical AFL. (Circulation. 1999;99: ) Key Words: atrial flutter atrium electrophysiology conduction Typical atrial flutter (AFL) is a macroreentrant rhythm propagating between both venae cavae and the tricuspid annulus (TA) For such a circuit to exist, a posterior electrical barrier is needed to prevent a short-circuiting between the posterior wall (PW) and lateral wall (LW). 11,12 In humans, this barrier is related to the crista terminalis (CT). 13 The fibers of this structure run longitudinally between the LW and PW in a high-to-low pattern. This structure, which is the fastest pathway in the lateral free wall in the craniocaudal direction, 14,15 creates a line of block in the transversal direction that determines the activation pattern of the LW during AFL. Although the CT is an electrical barrier during AFL, there are no data about the conduction properties of this structure after sinus rhythm has been restored, namely, if the conduction block is functional or fixed. The recording of double electrograms is a marker of areas of conduction block, but they are observed only during reentry or when the activation wave front is perpendicular to the line of block ,16,17 Thus, to determine whether the CT is a fixed or functional line of block, pacing was performed on both sides of the CT at several rates. The AFL circuit is critically dependent on conduction through the cavotricuspid isthmus (CTI), and this region is the target of transcatheter ablation procedures. 18,19 Because detection of CTI conduction block is based on changes in the activation patterns of the LW and interatrial septum during low LW and coronary sinus ostium (CSO) pacing, 20,21 an additional purpose of the present study was to determine the influence of CT conduction properties on activation patterns of the right atrium during pacing and therefore on CTI conduction-block evaluation. Methods Population The study group consisted of 22 patients with typical AFL, defined by an inverted sawtooth pattern in the inferior ECG leads and a Received October 23, 1998; revision received March 11, 1999; accepted March 23, From the Department of Cardiology, Hospital General Universitario Gregorio Marañón (A.A., J.A., J.V., J.L.M.S., N.P.-C., S.G., M.O., J.L.D.), Madrid, Spain, and Hospital de Basurto (J.M.A., J.M.O.), Bilbao, Spain. Correspondence to Angel Arenal, Laboratorio de Electrofisiología, Departamento de Cardiología, Hospital General Universitario Gregorio Marañón, C/Dr Esquerdo 46, Madrid, Spain. arenal@doymanet.es 1999 American Heart Association, Inc. Circulation is available at

2 2772 Conduction Block of the Crista Terminalis TABLE 1. Clinical Characteristics Patient Age, y/sex Cardiac Disease Clinical Arrhythmia Previous Antiarrhythmic Drugs 1 53/F AFL Fle 2 65/M HTN AFL Amio, Fle, Q 3 61/F CAD AFL -blockers 4 72/M AFL/AF Amio, Dig 5 66/M CAD AFL -blockers 6 61/M CAD AFL -blockers 7 54/M HTN AFL/AF Dig 8 66/M AFL Pro, Fle 9 69/M CAD AFL Amio, Dig 10 67/M CAD AFL S, V 11 81/M CAD AFL/AF Amio 12 68/M HTN/COPD AFL Amio 13 66/M CAD AFL -blockers 14 49/M AFL Pro 15 49/M HTN AFL 16 48/M CAD AFL Dig, -blockers 17 54/M AFL -blockers 18 46/M VD AFL/AF Amio 19 58/M COPD AFL Dig 20 60/M CAD AFL/AF -blockers 21 57/F AFL/AF Fle 22 67/M AFL Dig, -blockers Fle indicates flecainide; HTN, hypertension; Amio, amiodarone; Q, quinidine; CAD, coronary artery disease; Dig, digoxin, Pro, propafenone; S, sotalol; V, verapamil; COPD, chronic obstructive pulmonary disease; and VD, valvular disease. regular atrial rate 240 bpm in the absence of antiarrhythmic drugs, referred for radiofrequency ablation of the CTI (Table 1). At the time of the study, patients 4 and 12 were taking amiodarone, and patient 21 was taking flecainide; the remaining patients were not taking any class I or III antiarrhythmic drugs. Electrophysiological Testing: Electrical Stimulation and Recordings Studies were performed with patients in a nonsedated and postabsorptive state; written consent was obtained from each patient. Intracardiac recordings, which were filtered between 30 and 500 Hz with a gain amplification between 0.5 and 0.1 mv/cm, were displayed simultaneously with 1 ECG lead (II or avf) on a 12-channel recorder (Midas, Hellige Biomedical) at paper speeds of 100 and 200 mm/s. Atrial stimulation was performed with a programmable stimulator (UHS-20 BiotroniK) set to deliver rectangular pulses of 1-ms duration at twice the diastolic threshold. Figure 1 shows the fluoroscopic appearance of the catheter arrangement. A deflectable halo catheter (2-mm interelectrode distance, 10-mm interbipole distance; Webster Laboratories) was placed around the TA to obtain the right atrium activation sequence during flutter and the catheter ablation procedure; the distal pair was located close to the ablation line at approximately the 6-o clock position of the TA. A quadripolar catheter was placed at the CSO to test conduction between the septal and lateral walls through the CTI. A third quadripolar deflectable-tip catheter was used for radiofrequency application. This catheter was also used for CT identification by searching the double electrogram during flutter between the LW and PW. In 10 patients, a 20-pole deflectable catheter (Crista Catheter, Cordis), spacing mm, was placed at or in the proximity of the CT, and the distal electrode was next to the inferior vena cava. Radiofrequency Ablation After mapping and entrainment techniques characterized the arrhythmia as AFL, if sinus rhythm could be restored, pacing at both sides of the CTI was performed. Incremental pacing trains (from 600 ms to 2-to-1 atrial capture) were delivered at the CSO and from the distal pair of the halo catheters to establish right atrium activation patterns before radiofrequency ablation. A quadripolar deflectable-tip catheter was used for radiofrequency application. Radiofrequency ablation was performed with current generated by a conventional 500-kHz radiofrequency energy source (EPT-1000, EP Technologies or Atakr, Medtronic Cardiorhythm), delivered from the 8- or 4-mm tip of a steerable mapping catheter (Blazer T, EP Technologies and Marinr, Medtronic Cardiorhythm) to a left subscapular chest wall patch. Ablation was anatomically guided and performed during AFL or sinus Figure 1. Right and left anterior oblique fluoroscopic projections, respectively, showing catheter positioning. Halo catheter spans the TA from interatrial septum to CTI; distal pair is located at 6-o clock position. Crista catheter is placed between LW and PW, along area from which double electrograms were recorded. Distal electrode is close to inferior vena cava. A quadripolar catheter is at the CSO, and ablation (Abl) catheter shows location of ablation line.

3 Arenal et al June 1, TABLE 2. Electrophysiological Results Pacing Site LW PW CTIB Patient FCL, ms CT PB, ms CT CB, ms CT PB, ms CT CB, ms CCWB, ms CWB, ms ND ND ND ND ND ND ND ND ND Mean SD * * CTIB indicates CTI block; FCL, flutter cycle length; PB, partial block; CB, complete block; CCWB, counterclockwise block; CWB, clockwise block; ND, not determined. *P 0.01; P rhythm. Linear lesions were produced in the CTI, and the ablation catheter was progressively withdrawn under fluoroscopic guidance during radiofrequency energy delivery from the TA to the inferior vena cava (pulse duration between 90 and 120 seconds). Linear lesions were produced in the CTI in an attempt to achieve bidirectional conduction block between the LW and interatrial septum. To test the appearance of CTI block, the previously mentioned pacing protocol at both sides of the CTI was repeated after each ablation line was completed. CTI block was presumed to be present when the activation pattern of the side opposite to the pacing site was completely craniocaudal in nature. CT Location CT location was defined by recording double electrograms during flutter in the union of the LW and PW. Frames of the right and left anterior oblique fluoroscopic projections obtained during mapping were used to determine the location of the CT after sinus rhythm was restored. Evaluation of Conduction Across the CT After the radiofrequency ablation procedure was finished, in cases in which the Crista catheter was not used to record CT electrograms, the halo catheter was rotated counterclockwise and then pulled back to achieve close contact of the distal electrodes with the atrial wall at the proximity of the CT, so that 5 or 6 bipolar electrograms were recorded along the CT from the high to the low right atrium, which was done to maintain the distal electrode next to the inferior vena cava. After a stable position of the catheter was achieved, we proceeded to pace on both sides of the CT. To obtain activation wave fronts perpendicular to the CT, we selected sites from which the activation time differences among the bipolar electrograms recorded from the low to the high atrium were the shortest possible at the slowest rate. When it was difficult to locate a stable position at the PW, pacing was performed at the CSO. The second pacing site was at the low lateral right atrium. Multiple 10-second-duration synchronized trains of rapid atrial pacing at a constant rate were delivered at decremental cycle lengths from 600 ms until 2-to-1 atrial capture occurred. Definitions Double electrograms are defined as 2 discrete deflections separated by an isoelectric interval. Focal transversal conduction block in the CT was recognized by the recording of double electrograms at 1 site during pacing at the LW and PW. Complete transversal conduction block along the CT was detected by the appearance of double electrograms at all recording sites and a change in the activation sequence, with the development of a craniocaudal activation sequence at the opposite side of the pacing site. Recording of double electrograms at the highest portion of the CT was not necessary for the assumption of complete block, because this area was considered the turning point of the activation wave front. Constant clockwise block of the CTI was defined by observation of a completely descending activation of the LW during pacing from the CSO at 600 ms.

4 2774 Conduction Block of the Crista Terminalis Rate-dependent counterclockwise block of the CTI was presumed to occur when the CSO electrogram was activated simultaneous with or before the high interatrial septum and His bundle area during pacing at 600 ms, but later at a shorter cycle length. Conduction interval was defined as the interval between the CSO and the low LW (poles 1 and 2 of the halo catheter) during pacing from both sites at the shortest cycle length before CT block and at the pacing cycle length that provoked complete CT block. Statistical Analysis Values are expressed as mean SD. Statistical comparisons for 2 groups were performed with the Student s t test or the signed rank test. A value of P 0.05 (2-tailed) was considered significant. Figure 2. Records obtained in patient 12, in whom bidirectional transversal block in the CT was observed at the slowest pacing rate, suggesting fixed conduction block. Left, During pacing at 600 ms from PW, double electrograms were recorded from CT2 to CT 5. Second component of these electrograms showed craniocaudal activation, suggesting complete block of the CT. Right, During pacing at 600 ms from LW, double electrograms were also observed from CT 2 to CT 5. HRA indicates high right atrium; IS, interatrial septum. Rate-dependent clockwise block of the CTI was defined by the observation of 2 activation wave fronts, one ascending and the other descending, in the LW during pacing from the CSO at 600 ms, but only a single descending activation of the LW during pacing at shorter cycle lengths. Constant counterclockwise block of the CTI was presumed to occur when the CSO electrogram was activated after the high interatrial septum and His bundle area during pacing at 600 ms from the low lateral right atrium. Results Recordings obtained along the right atrium during flutter showed activation propagating in a counterclockwise direction in all patients. The flutter cycle length of this group was ms. After radiofrequency ablation, constant bidirectional CTI block was observed in 17 patients, whereas 3 patients showed clockwise rate-dependent block and 2 showed both counterclockwise and clockwise rate-dependent block. During sinus rhythm, neither double nor fragmented electrograms were recorded in the CT area. (See Table 2.) Conduction Across the CT Although complete transversal conduction block was observed in all patients from 1 pacing site, it was observed at the longest possible pacing cycle length in only a minority, suggesting the presence of fixed conduction block (patients 2, 11, 12, and 20 during pacing from the PW/CSO and patients 12 and 20 during pacing from the LW; Figure 2). In the remaining patients, the block was rate dependent (Figure 3). As shown in Table 2, both partial and complete block were achieved at longer pacing cycle lengths from the PW/CSO Figure 3. One ECG surface lead and 8 intracardiac electrograms recorded during sinus rhythm and pacing at 250 and 200 ms from PW. CT1 to CT6 are electrograms recorded from high to low right atrium in CT area. All CT electrograms were normal in amplitude and duration during sinus rhythm. During pacing at 250 ms, double electrograms appeared only at CT6; during pacing at 200 ms, double electrograms were clearly differentiated from CT3 to CT6. First components of electrograms at these sites were activated simultaneously; second components of electrograms (recorded at other side of CT) show a craniocaudal activation pattern, consistent with complete block of CT. Figure 4. Same patient as in Figure 3. Electrograms recorded during pacing from LW. With pacing at 250 ms, a 2-component electrogram appears only at CT6; during pacing at 200 ms, double electrograms are only separated at CT3. Remaining recordings also show double electrograms, but not separated. This is consistent with conduction block present at superior part of CT, whereas in the lower part there is only slowing of conduction.

5 Arenal et al June 1, Figure 5. Recording obtained in a patient with a seemingly rate-dependent clockwise conduction block at CTI during pacing from CSO. Four electrograms are recorded at LW with halo catheter: LW1 is recorded with distal pair at a site close to ablation line and LW4 at 9 o clock position. Another 5 electrograms are recorded along CT. CT5 is recorded close to the inferior vena cava. A, During pacing from CSO at 300 ms, activation sequence in LW is caudocranial, consistent with clockwise conduction in CTI, and no double electrograms were recorded at CT. B, Pacing at 250 ms: when double electrograms start to appear at CT, conduction interval to LW4 increases. C, Pacing at 240 ms: when double electrograms are clearly observed at CT, conduction interval to LW3 and LW4 increases. D, During pacing at 230 ms, double electrograms are recorded all along the CT with craniocaudal activation at opposite side to pacing site, suggesting complete block at CT. At that time, conduction interval to LW4 increases further, resulting in a change in activation pattern of LW compatible with block at CTI. This figure shows temporal and frequency relationship between appearance of block at CT and a change in activation pattern of LW. In this case, conduction across CT may mimic conduction across CTI. than from the LW. In 3 patients (patients 1, 6, and 13), complete block was not achieved with the shortest pacing cycle length from the LW (Figure 4), which suggests pacing site dependent conduction block. CT Conduction and CTI Block Assessment Constant clockwise CTI block was observed in 4 and 13 patients with fixed and rate-dependent CT conduction block, respectively. In 4 of the 5 cases in which CTI block was rate dependent (patients 15, 16, 18, and 19), the block appeared at the same cycle length as at the CT (Figure 5). In the remaining case (patient 14), CTI block appeared at a longer cycle length than for CT block (Figure 6). Constant counterclockwise CTI block was present in all but 2 patients, who had CT rate-dependent block. In 11 patients with rate-dependent CT conduction block, we could measure the conduction interval between the CSO and the low LW before and after the appearance of complete block at the CT (Table 3). It was during clockwise CTI block testing that the conduction interval increment was signifi-

6 2776 Conduction Block of the Crista Terminalis TABLE 3. Electrophysiological Results LW Pacing PW/CSO Pacing Figure 6. Recording obtained in a patient with true rate-dependent clockwise conduction block at CTI during pacing from CSO. A, During pacing from CSO at 500 ms, activation sequence in low LW is caudocranial, consistent with clockwise conduction in CTI, and although double electrograms are recorded at CT, activation sequence at lateral aspect of CT is not consistent with complete transversal block. B, Although during pacing at 400 ms craniocaudal activation pattern at LW suggests clockwise CTI block, CT electrograms remain unchanged. C, During pacing at 250 ms, double electrograms are recorded all along the CT, with craniocaudal activation at opposite side to pacing site, suggesting complete block at CT. This figure shows a change in activation pattern of LW preceding induction of complete block at CT. CCW-CTIB, CTCB, CWCTIB, CTCB, Patient ms ms CI, ms ms ms CI, ms 6 C 200 C C C C C C RD (400) RD (250) RD (250) C RD (230) C C C RD (230) RD (250) RD (350) C C C C CCW indicates counterclockwise; CTIB, CTI block; CTCB, complete block; CI, conduction interval difference during pacing before and after CT block; and CW, clockwise. cantly greater (67 39 versus 4 4 ms; P 0.02) in patients with rate-dependent block (Figure 5) than in patients with constant CTI block (Figure 7). Discussion The presence of a line of block along the CT is presumed to be essential for AFL to occur. A match between wavelength and circuit size is achieved if this line of block prevents the short circuit between the PW and LW, circumscribing the macroreentry around the tricuspid ring. We have shown that CT transversal conduction block in patients with typical AFL is rate and pacing-site dependent, with the exception that in some patients, complete block did not appear during pacing from the LW. We have also shown that conduction properties are not uniform along the CT; in most patients, partial block occurs at very different pacing cycle lengths than complete block. Complete bidirectional block at the CT was usually observed only at short cycle lengths. This observation may be the reason for the narrow limits of the flutter cycle length. Reentry would remain stable if conduction time along the macroreentrant circuit was shorter than the cycle length at which CT block developed. Differences Between PW and LW Rate-Dependent Transversal Conduction Block We observed that conduction block in the CT was achieved at slower pacing rates from the PW than from the LW, where only partial block was observed in some cases. Differences in cell arrangement between the smooth and trabeculated atrial walls may produce a different electrical input in the CT and consequently a different rate-dependent block. This characteristic could explain the greater incidence of counterclockwise rotation in spontaneous AFL. According to CT conduction characteristics, AFL should be more easily induced by atrial arrhythmias arising in the PW or the left atrium than by those arising from the high and lateral right atrium, which would require a shorter cycle length to cause complete conduction block. Because the direction of rotation of atrial

7 Arenal et al June 1, Figure 7. ECG surface lead and 12 intracardiac electrograms recorded during CSO pacing. AbC is ablation catheter. A, During pacing at 400 ms, no double or fragmented electrograms are recorded except at CT1. LW activation pattern is craniocaudal, and conduction interval to LW5 is 120 ms. B, During pacing at 300 ms, appearance of double electrograms at all CT recordings and activation pattern suggest CT complete block. Nevertheless, activation sequence at LW remains unchanged, and conduction interval is only 10 ms longer. Activation sequence of LW and conduction interval to low LW are independent whether conduction across CT is still present or not. fibrillation (AF) may be dependent on the site of induction (pacing from the PW induces counterclockwise flutter, and pacing from the lateral right atrium induces clockwise flutter 22,23 ), counterclockwise flutter should occur more frequently. In the same sense, atrial arrhythmias with coupling intervals or cycle lengths that provoke conduction block only from the PW to the LW will induce stable flutter if propagation is counterclockwise. If conduction block is only present from the PW to the LW and conduction is possible in the opposite direction, the only way for the CT to act as a line of bidirectional conduction block is if the postactivation refractoriness of the CT, just before the line of block, exceeds the time required to surround the line of block. This will depend on the conduction time of the activation wave front around the CT. Therefore, when activation is counterclockwise, propagation between both sides of the CT through the high right atrium will have a short delay, and the CT will function as a barrier in both directions. If activation of the opposite side of the CT is delayed (ie, clockwise propagation through the previously mentioned slow-conducting area located in the low atrium), the barrier of refractoriness could end, because there is enough time to recover and conduction to the other side of the CT is possible at some level of the CT, thus creating smaller and more unstable circuits. This hypothesis could explain why counterclockwise right atrium circus movement is more commonly observed than clockwise movement in type 1 AFL. Nevertheless, some other determinants are presumably involved, because AFL in transplanted hearts is usually counterclockwise despite the lack of any role for the CT in the flutter mechanism. 24 Influence of CT Rate-Dependent Conduction Block on the Evaluation of CTI Block The CTI is commonly used as the target of radiofrequency ablation. 18,19 Achievement of constant block at this isthmus is the best marker of success. 20,21 Nevertheless, rate-dependent block does not imply recurrence during follow-up. 21,25 A complete craniocaudal activation pattern of the opposite wall to the pacing site (CSO and low LW) is consistent with CTI block. Theoretically, to observe this activation pattern, a posterior electrical barrier should be present to prevent short-circuiting between the PW and LW. Because this line of block is usually functional, a change would be expected in the activation pattern and conduction interval when functional block appears at the CT. This hypothesis is consistent with the observation of rate-dependent CTI block, as was seen in the cases in which CT and CTI blocks appeared simultaneously. In these patients, a short-circuiting of the low LW at lower rates could mimic conduction along the CTI, thus precluding recognition of CTI block. Permanent CTI block was observed, as expected, in all patients in whom we could demonstrate fixed block at the CT, but it was also seen in the majority of cases of rate-dependent block. This surprising observation suggests that activation of the LW, at least close to the TA, is in some cases independent of the conduction state of the CT. In these patients, the conduction interval from the CSO to the low LW is not modified by the appearance of block at the CT, probably because the conduction velocity is faster along the longitudi-

8 2778 Conduction Block of the Crista Terminalis nal fibers surrounding the TA than across the CT. In these cases, the block along the CT may be essential for the initiation of AFL but not for its perpetuation. Study Limitations There are several limitations to this study. The CT was identified by the recording of double electrograms during flutter; we did not use intravascular echocardiography. Nevertheless, in all cases, we were able to record double electrograms at a particular location between the LW and PW that were stable throughout the entire pacing protocol; therefore, we can assume there is constant location for this line of block, and the electrophysiological implications are identical whether the line of block is at or close to the CT. Another limitation is that we obtained data from only 1 selected pacing site on each side of the CT, and this paced activation wave front might be different from that observed during flutter. Therefore, rate dependency may be different during flutter and pacing. Clinical Implications The production of bidirectional block between the lateral and septal walls in the CTI is currently being used as the end point of radiofrequency catheter ablation of AFL. 20,21 Nevertheless, in patients in whom the CT does not act as a barrier between the LW and PW at the flutter cycle length, a clockwise rotating circuit may not be stable because of its absence. Thus, in these patients, it is conceivable that bidirectional block is unnecessary and that unidirectional block between the lateral and septal walls may suffice to treat AF by catheter ablation. In cases with rate-dependent CTI block, it is important to determine the relationship with the CT ratedependent block to avoid the possibility that conduction across the CT could mimic permeability across the CTI. References 1. Lewis T, Feil HS, Stroud WD. Observations upon flutter and fibrillation, II: the nature of auricular flutter. Heart. 1920;7: Rosenblueth A, Garcia-Ramos J. Studies on flutter and fibrillation, II: the influence of artificial obstacles on experimental auricular flutter. Am Heart J. 1947;33: Kimura E, Kato S, Murao S, Ajisaka H, Koyama S, Omiya Z. Experimental studies on the mechanism of the auricular flutter. Tohoku J Exp Med. 1954;60: Hayden WG, Hurley EJ, Rytand DAS. The mechanism of canine atrial flutter. Circ Res. 1967;20: Inoue H, Matsuo H, Takayanagi K, Murao S. Clinical and experimental studies of the effects of atrial extrastimulation and rapid pacing on atrial flutter cycle: evidence of macro-reentry with an excitable gap. Am J Cardiol. 1981;48: Cosio FG, Arribas F, Palacios J, Tascon J, Lopez-Gil M. Fragmented electrograms and continuous electrical activity in atrial flutter. Am J Cardiol. 1986;57: Olshanky B, Okumura K, Hess PG, Waldo AL. Demonstration of an area of slow conduction in human atrial flutter. J Am Coll Cardiol. 1990;16: Arenal A, Almendral J, San Roman D, Delcan JL, Josephson ME. Frequency and implications of resetting and entrainment with right atrial stimulation in atrial flutter. Am J Cardiol. 1992;70: Kalman JM, Olgin JE, Saxon LA, Fisher WG, Lee RJ, Lesh MD. Activation and entrainment mapping defines the tricuspid annulus as the anterior barrier in typical atrial flutter. Circulation. 1996;94: Almendral JM, Arenal A. Electrophysiology of human atrial flutter. In: Josephson ME, Wellens HJJ, eds. Tachycardias: Mechanism and Management. Mount Kisco, NY: Futura Publishing; 1993: Cosio FG, Arribas F, Barbero JM. Validation of double spike electrograms as markers of conduction delay or block in atrial flutter. Am J Cardiol. 1988;61: Olshansky B, Okumura K, Henthorn RW, Waldo AL. Characterization of double potential in human atrial flutter: studies during transient entrainment. J Am Coll Cardiol. 1990;15: Olgin JE, Kalman JM, Fitzpatrick AP, Lesh MD. Role of right atrial endocardial structures as barriers to conduction during human type I atrial flutter: activation and entrainment mapping guided by intracardiac echocardiography. Circulation. 1995;92: Pressler ML, Münster PN, Huang X. Gap junction distribution in the heart: functional relevance. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology: From Cell to Bedside. Philadelphia, Pa: WB Saunders; 1995: Yamasita T, Inoue H, Nozaki A, Sugimoto T. Role of anatomic architecture in sustained atrial reentry and double potentials. Am Heart J. 1992;124: Feld G, Shahandeh-Rad F. Mechanism of double potentials recorded during sustained atrial flutter in the canine right atrial crush-injury model. Circulation. 1992;86: Nakagawa H, Lazzara R, Khastgir T, Beckman KJ, McClelland JH, Imai S, Pitha J, Becker AE, Arruda M, Gonzalez MD, Widman LE, Rome M, Neuhauser J, Wang X, Calame J, Goudeau MD, Jackman WM. Role of the tricuspid annulus and the eustachian valve/ridge on atrial flutter: relevance to catheter ablation of the septal isthmus and a new technique for rapid identification of ablation success. Circulation. 1996;94: Cosio FG, Lopez-Gil M, Giocolea A, Arribas F, Barroso JL. Radiofrequency ablation of the inferior vena cava-tricuspid valve isthmus in common atrial flutter. Am J Cardiol. 1993;71: Fischer B, Jaïs P, Shah DC, Chouairi S, Haïssaguerre M, Garrigue S, Poquet F, Gencel L, Clémenty J, Marcus FI. Radiofrequency catheter ablation of common atrial flutter in 200 patients. J Cardiovasc Electrophysiol. 1996;7: Poty H, Saoudi N, Aziz AA, Nari M, Letac B. Radiofrequency catheter ablation of type I atrial flutter: prediction of late success by electrophysiological criteria. Circulation. 1995;92: Cauchemez B, Haïssaguerre M, Fischer B, Thomas O, Clémenty J, Coumel P. Electrophysiological effects of catheter ablation of inferior vena cava tricuspid annulus isthmus in common atrial flutter. Circulation. 1996;93: Olgin JE, Kalman JM, Saxon LA, Lee RJ, Lesh MD. Mechanism of initiation of atrial flutter in humans: site of unidirectional block and direction of rotation. J Am Coll Cardiol. 1997;29: Tai CT, Chen SA, Chiang CE, Lee SH, Ueng KC, Wen ZC, Huang JL, Chen YJ, Yu WC, Feng AN, Chiou CW, Chang MS. Characterization of low atrial isthmus as the slow conduction zone and pharmacological target in typical atrial flutter. Circulation. 1997;96: Arenal A, Almendral J, Muñoz R, Villacastín J, Merino JL, Palomo J, Garcia-Robles JA, Peinado R, Delcán JL. Mechanism and location of atrial flutter in transplanted hearts: observations during transient entrainment from distant sites. J Am Coll Cardiol. 1997;30: Schwartzman D, Callans DJ, Gottlieb SD, Dillon SM, Movsowitz C, Marchlinski FE. Conduction block in the inferior vena caval-tricuspid valve isthmus: association with outcome of radiofrequency ablation of type I atrial flutter. J Am Coll Cardiol. 1996;28: